Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “scientific deep learning”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 235 records · Page 13
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Nanostructures for Electrical Energy Storage (NEES) (2020 Final Technical Report)

Nanostructures for Electrical Energy Storage (NEES, www.efrc.umd.edu) was an Energy Frontier Research Center supported by the DOE Office of Science, Basic Energy Sciences, from 8/1/2009 to 7/31/2020. Led by the University of Maryland, NEES enjoyed extensive collaborations with its funded partners, including two DOE Laboratories and six universities. The NEES vision has been to reveal a set of scientific insights and design principles that can underpin a next-generation electrical energy storage approach, building on advances in nanoscale science and technology to achieve simultaneous high power and high energy over extended charge/discharge cycling. The vision is motivated by the recognition that scaling into the nano regime opens the door to new physical phenomena and that the tools enlisted in nanoscale research provide major new opportunities for the synthesis not only of materials at molecular scale but for structures at nano scale and above. NEES has translated this vision into its research program based on two observations. First, while the behavior of ions and electrons in electrolytes and in electrode materials is crucial to electrical energy storage (or more appropriately electrochemical energy storage), it is the transport of ion and electron charge between different structural components of a storage device that ultimately determine its performance. With it well recognized that the choice of electrode materials typically constrain ion transport kinetics as well as maximum ion concentration, the search for better electrode materials has been a primary driver of battery research. At the same time the synthesis of electrodes is typically based on aggregation of particles with varying size, shape, and orientation in the electrode. Together with the presence of additional materials to impart electrical conductivity and cohesion to the composite electrode, change in electrode materials is necessarily accompanied by structural changes at the nano/micro scale that are difficult to categorize and manage. From the beginning, NEES’ vision has been to create and study simpler, highly controlled spatial arrangements of known materials as battery components (electrodes, current collectors, and electrolyte) and to understand how design and structure above the molecular scale determines the energy storage performance available from known materials. Second, advances in nanoscience dramatically expanded the portfolio of synthesis methods, structural motifs, and new phenomena available for research. Some of these gave rapid access to new building blocks at the deep nanoscale (e.g., carbon nanotubes grown by self-assembly, nanoscale arrays formed by electrochemical self-alignment, monolayer films controlled by self-limiting reaction). Such advances served as the enabler for the NEES vision to be pursued experimentally through study of 3D structures created and controlled at the nano, micro, and meso scales. Here, we use meso as in the BES MESO Report, implying not only intermediate or varying length scales, but very much the way behavior is influenced by other factors including aggregation of nanocomponents at different densities and spatial configurations, statistical variations in the aggregates, hierarchical architectures in which they can be assembled, or local 3D configurations that result from the architectures. Over its life cycle, NEES has pursued two overarching goals: (1) to understand the scientific fundamentals of electrochemical storage from the nanoscale to the mesoscale; and (2) to create and learn from innovative, controlled, heterogeneous nanostructures, where such nanostructures can enable the first goal and serve as models for future paradigms in energy storage. Specific goals have included: Synthesize heterogeneous nanostructures comprised of multiple materials arranged in controlled fashion and characterize their behavior; Demonstrate and elucidate design principles for achieving simultaneous high power and high energy; Develop materials processes which enable precision control of thin layers and 3D structures; Investigate the impact of artificial interphases on electrode stability during ion insertion/deinsertion; Create dense arrays of nanostructures to understand how the architecture of these assemblies, along with nanostructure design, influences energy storage behavior at the mesoscale; Identify and understand the consequences of nanoconfinement and local inhomogeneities in 3D mesoscale arrays; Develop and apply computational models to stimulate, guide and interpret experiments.

25 ENERGY STORAGE↗
DOE OSTI2023

CoolPINNs: A physics-informed neural network modeling of active cooling in vascular systems

Emerging technologies like hypersonic aircraft, space exploration vehicles, and batteries avail fluid circulation in embedded microvasculatures for efficient thermal regulation. Modeling is vital during the design and operational phases of these engineered systems. However, many challenges exist in developing a modeling framework. What is lacking is an accurate framework that (i) captures sharp jumps in the thermal flux across complex vasculature layouts, (ii) deals with oblique derivatives (involving tangential and normal components), (iii) handles nonlinearity because of radiative heat transfer, (iv) provides a high-speed forecast for real-time monitoring, and (v) facilitates robust inverse modeling. Here, this paper addresses these challenges by availing the power of physics-informed neural networks (PINNs). We develop a fast, reliable, and accurate Scientific Machine Learning (SciML) framework for vascular-based thermal regulation—called CoolPINNs: a PINNs-based modeling framework for active cooling. The proposed mesh-less framework elegantly overcomes all the mentioned challenges. The significance of the reported research is multi-fold. First, the framework is valuable for real-time monitoring of thermal regulatory systems because of rapid forecasting. Second, researchers can address complex thermoregulation designs since the approach is meshless. Finally, the framework facilitates systematic parameter identification and inverse modeling studies, perhaps the most significant utility of the current framework.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Multidimensional Distributional Neural Network Output Demonstrated in Super‐Resolution of Surface Wind Speed

Accurate quantification of uncertainty in neural network predictions remains a central challenge for scientific applications involving high-dimensional, correlated data. While existing methods capture either aleatoric or epistemic uncertainty, few offer closed-form, multidimensional distributions that preserve spatial correlation while remaining computationally tractable. In this work, we present a framework for training neural networks with a multidimensional Gaussian loss, generating a closed-form predictive distribution over outputs informed by non-identically distributed training data. Our approach captures aleatoric uncertainty by iteratively estimating the means and covariance matrices, and is demonstrated on a super-resolution example out-of-training-sample. We leverage a Fourier representation of the covariance matrix to stabilize network training and preserve spatial correlation. We introduce a novel regularization strategy—referred to as information sharing—that interpolates between image-specific and global covariance estimates, enabling convergence of the super-resolution downscaling network trained on image-specific distributional loss functions. This framework allows for efficient sampling, explicit correlation modeling, and extensions to more complex distribution families all without disrupting prediction performance. We demonstrate the method on a surface wind speed downscaling task and discuss its broader applicability to uncertainty-aware prediction in scientific models.

17 WIND ENERGY↗
DOE OSTI2023

Context-aware learning of hierarchies of low-fidelity models for multi-fidelity uncertainty quantification

Multi-fidelity Monte Carlo methods leverage low-fidelity and surrogate models for variance reduction to make tractable uncertainty quantification even when numerically simulating the physical systems of interest with high-fidelity models is computationally expensive. This work proposes a context-aware multi-fidelity Monte Carlo method that optimally balances the costs of training low-fidelity models with the costs of Monte Carlo sampling. It generalizes the previously developed context-aware bi-fidelity Monte Carlo method to hierarchies of multiple models and to more general types of low-fidelity models. When training low-fidelity models, the proposed approach takes into account the context in which the learned low-fidelity models will be used, namely for variance reduction in Monte Carlo estimation, which allows it to find optimal trade-offs between training and sampling to minimize upper bounds of the mean-squared errors of the estimators for given computational budgets. This is in stark contrast to traditional surrogate modeling and model reduction techniques that construct low-fidelity models with the primary goal of approximating well the high-fidelity model outputs and typically ignore the context in which the learned models will be used in upstream tasks. Further, the proposed context-aware multi-fidelity Monte Carlo method applies to hierarchies of a wide range of types of low-fidelity models such as sparse-grid and deep-network models. Numerical experiments with the gyrokinetic simulation code Gene show speedups of up to two orders of magnitude compared to standard estimators when quantifying uncertainties in small-scale fluctuations in confined plasma in fusion reactors. This corresponds to a runtime reduction from 72 days to four hours on one node of the Lonestar6 supercomputer at the Texas Advanced Computing Center.

42 ENGINEERING↗
DOE OSTI2025

Self-supervised physics-informed generative networks for phase retrieval from a single X-ray hologram

X-ray phase contrast imaging significantly improves the visualization of structures with weak or uniform absorption, broadening its applications across a wide range of scientific disciplines. Propagation-based phase contrast is particularly suitable for time- or dose-critical in vivo/in situ/operando (tomography) experiments because it requires only a single intensity measurement. However, the phase information of the wave field is lost during the measurement and must be recovered. Conventional algebraic and iterative methods often rely on specific approximations or boundary conditions that may not be met by many samples or experimental setups. In addition, they require manual tuning of reconstruction parameters by experts, making them less adaptable for complex or variable conditions. Here we present a self-learning approach for solving the inverse problem of phase retrieval in the near-field regime of Fresnel theory using a single intensity measurement (hologram). A physics-informed generative adversarial network is employed to reconstruct both the phase and absorbance of the unpropagated wave field in the sample plane from a single hologram. Unlike most state-of-the-art deep learning approaches for phase retrieval, our approach does not require paired, unpaired, or simulated training data. This significantly broadens the applicability of our approach, as acquiring or generating suitable training data remains a major challenge due to the wide variability in sample types and experimental configurations. The algorithm demonstrates robust and consistent performance across diverse imaging conditions and sample types, delivering quantitative, high-quality reconstructions for both simulated data and experimental datasets acquired at beamline P05 at PETRA III (DESY, Hamburg), operated by Helmholtz-Zentrum Hereon. Furthermore, it enables the simultaneous retrieval of both phase and absorption information.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Self-Driving Telescopes: Autonomous Scheduling of Astronomical Observation Campaigns with Offline Reinforcement Learning

Modern astronomical experiments are designed to achieve multiple scientific goals, from studies of galaxy evolution to cosmic acceleration. These goals require data of many different classes of night-sky objects, each of which has a particular set of observational needs. These observational needs are typically in strong competition with one another. This poses a challenging multi-objective optimization problem that remains unsolved. The effectiveness of Reinforcement Learning (RL) as a valuable paradigm for training autonomous systems has been well-demonstrated, and it may provide the basis for self-driving telescopes capable of optimizing the scheduling for astronomy campaigns. Simulated datasets containing examples of interactions between a telescope and a discrete set of sky locations on the celestial sphere can be used to train an RL model to sequentially gather data from these several locations to maximize a cumulative reward as a measure of the quality of the data gathered. We use simulated data to test and compare multiple implementations of a Deep Q-Network (DQN) for the task of optimizing the schedule of observations from the Stone Edge Observatory (SEO). We combine multiple improvements on the DQN and adjustments to the dataset, showing that DQNs can achieve an average reward of 87%+-6% of the maximum achievable reward in each state on the test set. This is the first comparison of offline RL algorithms for a particular astronomical challenge and the first open-source framework for performing such a comparison and assessment task.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

The LSST AGN Data Challenge: Selection Methods

Abstract Development of the Rubin Observatory Legacy Survey of Space and Time (LSST) includes a series of Data Challenges (DCs) arranged by various LSST Scientific Collaborations that are taking place during the project's preoperational phase. The AGN Science Collaboration Data Challenge (AGNSC-DC) is a partial prototype of the expected LSST data on active galactic nuclei (AGNs), aimed at validating machine learning approaches for AGN selection and characterization in large surveys like LSST. The AGNSC-DC took place in 2021, focusing on accuracy, robustness, and scalability. The training and the blinded data sets were constructed to mimic the future LSST release catalogs using the data from the Sloan Digital Sky Survey Stripe 82 region and the XMM-Newton Large Scale Structure Survey region. Data features were divided into astrometry, photometry, color, morphology, redshift, and class label with the addition of variability features and images. We present the results of four submitted solutions to DCs using both classical and machine learning methods. We systematically test the performance of supervised models (support vector machine, random forest, extreme gradient boosting, artificial neural network, convolutional neural network) and unsupervised ones (deep embedding clustering) when applied to the problem of classifying/clustering sources as stars, galaxies, or AGNs. We obtained classification accuracy of 97.5% for supervised models and clustering accuracy of 96.0% for unsupervised ones and 95.0% with a classic approach for a blinded data set. We find that variability features significantly improve the accuracy of the trained models, and correlation analysis among different bands enables a fast and inexpensive first-order selection of quasar candidates.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2024

RandONets: Shallow networks with random projections for learning linear and nonlinear operators

Deep neural networks have been extensively used for the solution of both the forward and the inverse problem for dynamical systems. However, their implementation necessitates optimizing a high-dimensional space of parameters and hyperparameters. This fact, along with the requirement of substantial computational resources, pose a barrier to achieving high numerical accuracy, but also interpretability. Here, to address the above challenges, we present Random Projection-based Operator Networks (RandONets): shallow networks with random projections and tailor-made numerical analysis methods that learn accurately and fast linear and nonlinear operators. Building on previous works, we prove that RandOnets are universal approximators of linear and nonlinear operators. Due to their simplicity, RandONets provide a one-step transformation of the input space, facilitating interpretability. For the evaluation of their performance, we focus on operators of PDEs. We show, that RandONets outperform by several orders of magnitude, both in terms of numerical approximation accuracy and computational cost, the “vanilla” DeepONets. Hence, we believe that our method will trigger further developments in the field of scientific machine learning, for the development of new ‘’light”schemes that will provide high accuracy while reducing dramatically the computational cost. A MATLAB toolbox for RandONets, including demos, is available on GitHub at https://github.com/GianlucaFabiani/RandONets.

Interpretable machine learning↗
DOE OSTI2025

Tools for unbinned unfolding

Machine learning has enabled differential cross section measurements that are not discretized. Going beyond the traditional histogram-based paradigm, these unbinned unfolding methods are rapidly being integrated into experimental workflows. Here, in order to enable widespread adaptation and standardization, we develop methods, benchmarks, and software for unbinned unfolding. For methodology, we demonstrate the utility of boosted decision trees for unfolding with a relatively small number of high-level features. This complements state-of-the-art deep learning models capable of unfolding the full phase space. To benchmark unbinned unfolding methods, we develop an extension of existing dataset to include acceptance effects, a necessary challenge for real measurements. Additionally, we directly compare binned and unbinned methods using discretized inputs for the latter in order to control for the binning itself. Lastly, we have assembled two software packages for the OmniFold unbinned unfolding method that should serve as the starting point for any future analyses using this technique. One package is based on the widely-used RooUnfold framework and the other is a standalone package available through the Python Package Index (PyPI).

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Deep transfer operator learning for partial differential equations under conditional shift

Transfer learning enables the transfer of knowledge gained while learning to perform one task (source) to a related but different task (target), hence addressing the expense of data acquisition and labelling, potential computational power limitations and dataset distribution mismatches. Here, we propose a new transfer learning framework for task-specific learning (functional regression in partial differential equations) under conditional shift based on the deep operator network (DeepONet). Task-specific operator learning is accomplished by fine-tuning task-specific layers of the target DeepONet using a hybrid loss function that allows for the matching of individual target samples while also preserving the global properties of the conditional distribution of the target data. Inspired by conditional embedding operator theory, we minimize the statistical distance between labelled target data and the surrogate prediction on unlabelled target data by embedding conditional distributions onto a reproducing kernel Hilbert space. We demonstrate the advantages of our approach for various transfer learning scenarios involving nonlinear partial differential equations under diverse conditions due to shifts in the geometric domain and model dynamics. Our transfer learning framework enables fast and efficient learning of heterogeneous tasks despite considerable differences between the source and target domains.

42 ENGINEERING↗
DOE OSTI2025

Optimizing the optimizer for physics-informed neural networks and Kolmogorov-Arnold networks

Physics-Informed Neural Networks (PINNs) have revolutionized the computation of PDE solutions by integrating partial differential equations (PDEs) into the neural network’s training process as soft constraints, becoming an important component of the scientific machine learning (SciML) ecosystem. More recently, physics-informed Kolmogorv-Arnold networks (PIKANs) have also shown to be effective and comparable in accuracy with PINNs. In their current implementation, both PINNs and PIKANs are mainly optimized using first-order methods like Adam, as well as quasi-Newton methods such as BFGS and its low-memory variant, L-BFGS. However, these optimizers often struggle with highly nonlinear and non-convex loss landscapes, leading to challenges such as slow convergence, local minima entrapment, and (non)degenerate saddle points. In this study, we investigate the performance of Self- Scaled BFGS (SSBFGS), Self-Scaled Broyden (SSBroyden) methods and other advanced quasi-Newton schemes, including BFGS and L-BFGS with different line search strategies. These methods dynamically rescale updates based on historical gradient information, thus enhancing training efficiency and accuracy. We systematically compare these optimizers – using both PINNs and PIKANs – on key challenging PDEs, including the Burgers, Allen-Cahn, Kuramoto-Sivashinsky, Ginzburg-Landau, and Stokes equations. Additionally, we evaluate the performance of SSBFGS and SSBroyden for Deep Operator Network (DeepONet) architectures, demonstrating their effectiveness for data-driven operator learning. Our findings provide state-of-the-art results with orders-of-magnitude accuracy improvements without the use of adaptive weights or any other enhancements typically employed in PINNs. More broadly, our work reveal insights into the effectiveness of quasi-Newton optimization strategies in significantly improving the convergence and accurate generalization of PINNs and PIKANs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

COVID-19 dynamics across the US: A deep learning study of human mobility and social behavior

This paper presents a deep learning framework for epidemiology system identification from noisy and sparse observations with quantified uncertainty. The proposed approach employs an ensemble of deep neural networks to infer the time-dependent reproduction number of an infectious disease by formulating a tensor-based multi-step loss function that allows us to efficiently calibrate the model on multiple observed trajectories. The method is applied to a mobility and social behavior-based SEIR model of COVID-19 spread. The model is trained on Google and Unacast mobility data spanning a period of 66 days, and is able to yield accurate future forecasts of COVID-19 spread in 203 US counties within a time-window of 15 days. Interestingly, a sensitivity analysis that assesses the importance of different mobility and social behavior parameters reveals that attendance of close places, including workplaces, residential, and retail and recreational locations, has the largest impact on the effective reproduction number. Furthermore, the model enables us to rapidly probe and quantify the effects of government interventions, such as lock-down and re-opening strategies. Taken together, the proposed framework provides a robust workflow for data-driven epidemiology model discovery under uncertainty and produces probabilistic forecasts for the evolution of a pandemic that can judiciously provide information for policy and decision making. All codes and data accompanying this manuscript are available at https://github.com/PredictiveIntelligenceLab/DeepCOVID19.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

STSR-INR: Spatiotemporal super-resolution for multivariate time-varying volumetric data via implicit neural representation

Implicit neural representation (INR) has surfaced as a promising direction for solving different scientific visualization tasks due to its continuous representation and flexible input and output settings. We present STSR-INR, an INR solution for generating simultaneous spatiotemporal super-resolution for multivariate time-varying volumetric data. Inheriting the benefits of the INR-based approach, STSR-INR supports unsupervised learning and permits data upscaling with arbitrary spatial and temporal scale factors. Unlike existing GAN- or INR-based super-resolution methods, STSR-INR focuses on tackling variables or ensembles and enabling joint training across datasets of various spatiotemporal resolutions. Here we achieve this capability via a variable embedding scheme that learns latent vectors for different variables. In conjunction with a modulated structure in the network design, we employ a variational auto-decoder to optimize the learnable latent vectors to enable latent-space interpolation. To combat the slow training of INR, we leverage a multi-head strategy to improve training and inference speed with significant speedup. We demonstrate the effectiveness of STSR-INR with multiple scalar field datasets and compare it with conventional tricubic+linear interpolation and state-of-the-art deep-learning-based solutions (STNet and CoordNet).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Final DOE-ASR Report for the Project “Using LASSO to bridge the gap between model and observations and to learn about atmospheric convection”

Atmospheric convection spans a wide range of spatial and temporal scales and involves complex interactions with the surrounding dynamic and thermodynamic environment, particularly over tropical continental regions. These processes remain a major source of uncertainty in weather and climate models, including persistent biases in the diurnal cycle of convective precipitation that directly affect estimates of climate sensitivity. Addressing these challenges requires the combined use of high-resolution observations and cloud-resolving modeling frameworks. In this context, the DOE Atmospheric Radiation Measurement (ARM) program’s Large-Eddy Simulation ARM Symbiotic Simulation and Observation (LASSO) activity provides a powerful platform that pairs comprehensive observations with numerical simulations to enable process-level understanding of atmospheric convection. Within this context, this Research and Development Partnership Pilot (RDPP) project was designed to initiate and expand DOE ARM/ASR research capacity at minority-serving institutions, while advancing scientific understanding of convective processes over the Amazon rainforest. Consistent with the RDPP mission, the project emphasized partnership development, training, and workforce capacity building alongside exploratory research activities. On the scientific side, the project produced two peer-reviewed journal articles, and one manuscript currently under review (see list in section 3.1). Together, these studies combine long-term ARM observations and cloud-resolving and convection-permitting modeling to investigate the environmental controls on the shallow-to-deep convective transition during the Amazon wet season. The results demonstrate the central role of early-day moisture preconditioning and large-scale dynamical forcing in regulating isolated deep convection, provide mechanistic insight into convective evolution, and establish physically informed modeling frameworks for future sensitivity experiments. These scientific outcomes are described in sections 2.1 to 2.3 and were disseminated in 8 conference presentations (see section 3.2) and 5 invited talks (see section 3.3), reflecting broad engagement with our community. Equally important, the project achieved its RDPP capacity-building objectives (see section 2.4). A sustained research partnership was established among the University of Maryland, Baltimore County (UMBC), Morgan State University (MSU), and Howard University (HU), and extended to include collaboration with Pacific Northwest National Laboratory (PNNL). The project organized multiple multi-day training events focused on ARM data, LASSO simulations, and quantitative analysis methods, directly engaging students, postdoctoral researchers, and faculty across institutions. These activities broadened participation in ASR research and led to independent adoption of LASSO workflows by students beyond the immediate project team. Finally, the project successfully positioned the participating institutions to pursue future DOE research. Preliminary scientific results, coupled with strengthened partnerships and technical capacity, enabled the submission of follow-on proposals to DOE ASR funding opportunities. In this way, the project fulfilled the RDPP goal of seeding durable research capacity and laying the foundation for larger-scale, sustained engagement with DOE ARM and ASR programs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Accelerating Collective Communication in Data Parallel Training across Deep Learning Frameworks

This work develops new techniques within Horovod, a generic communication library supporting data parallel training across deep learning frameworks. In particular, we improve the Horovod control plane by implementing a new coordination scheme that takes advantage of the characteristics of the typical data parallel training paradigm, namely the repeated execution of collectives on the gradients of a fixed set of tensors. Using a caching strategy, we execute Horovod’s existing coordinator-worker logic only once during a typical training run, replacing it with a more efficient decentralized orchestration strategy using the cached data and a global intersection of a bitvector for the remaining training duration. Next, we introduce a feature for end users to explicitly group collective operations, enabling finer grained control over the communication buffer sizes. To evaluate our proposed strategies, we conduct experiments on a world-class supercomputer — Summit. We compare our proposals to Horovod’s original design and observe 2x performance improvement at a scale of 6000 GPUs; we also compare them against tf.distribute and torch.DDP and achieve 12% better and comparable performance, respectively, using up to 1536 GPUs; we compare our solution against BytePS in typical HPC settings and achieve about 20% better performance on a scale of 768 GPUs. Finally, we test our strategies on a scientific application (STEMDL) using up to 27,600 GPUs (the entire Summit) and show that we achieve a near-linear scaling of 0.93 with a sustained performance of 1.54 exaflops (with standard error +- 0.02) in FP16 precision.

Romero, Joshua↗
DOE OSTI2022

Interpretable AI forecasting for numerical relativity waveforms of quasicircular, spinning, nonprecessing binary black hole mergers

We present a deep-learning artificial intelligence model (AI) that is capable of learning and forecasting the late-inspiral, merger and ringdown of numerical relativity waveforms that describe quasicircular, spinning, nonprecessing binary black hole mergers. We used the NRHybSur3dq8 surrogate model to produce train, validation and test sets of ℓ = |m| = 2 waveforms that cover the parameter space of binary black hole mergers with mass ratios q ≤ 8 and individual spins |s$^{z}_ {(1,2)}$| ≤ 0.8. These waveforms cover the time range t ∊ [-5000 M, 130 M], where t = 0M marks the merger event, defined as the maximum value of the waveform amplitude. We harnessed the ThetaGPU supercomputer at the Argonne Leadership Computing Facility to train our AI model using a training set of 1.5 million waveforms. We used 16 NVIDIA DGX A100 nodes, each consisting of 8 NVIDIA A100 Tensor Core GPUs and 2 AMD Rome CPUs, to fully train our model within 3.5 h. Our findings show that artificial intelligence can accurately forecast the dynamical evolution of numerical relativity waveforms in the time range t ∊ [-100 M, 130 M]. Sampling a test set of 190,000 waveforms, we find that the average overlap between target and predicted waveforms is ≳99% over the entire parameter space under consideration. We also combined scientific visualization and accelerated computing to identify what components of our model take in knowledge from the early and late-time waveform evolution to accurately forecast the latter part of numerical relativity waveforms. This work aims to accelerate the creation of scalable, computationally efficient and interpretable artificial intelligence models for gravitational wave astrophysics

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Scalable training of graph convolutional neural networks for fast and accurate predictions of HOMO-LUMO gap in molecules

Abstract Graph Convolutional Neural Network (GCNN) is a popular class of deep learning (DL) models in material science to predict material properties from the graph representation of molecular structures. Training an accurate and comprehensive GCNN surrogate for molecular design requires large-scale graph datasets and is usually a time-consuming process. Recent advances in GPUs and distributed computing open a path to reduce the computational cost for GCNN training effectively. However, efficient utilization of high performance computing (HPC) resources for training requires simultaneously optimizing large-scale data management and scalable stochastic batched optimization techniques. In this work, we focus on building GCNN models on HPC systems to predict material properties of millions of molecules. We use HydraGNN, our in-house library for large-scale GCNN training, leveraging distributed data parallelism in PyTorch. We use ADIOS, a high-performance data management framework for efficient storage and reading of large molecular graph data. We perform parallel training on two open-source large-scale graph datasets to build a GCNN predictor for an important quantum property known as the HOMO-LUMO gap. We measure the scalability, accuracy, and convergence of our approach on two DOE supercomputers: the Summit supercomputer at the Oak Ridge Leadership Computing Facility (OLCF) and the Perlmutter system at the National Energy Research Scientific Computing Center (NERSC). We present our experimental results with HydraGNN showing (i) reduction of data loading time up to 4.2 times compared with a conventional method and (ii) linear scaling performance for training up to 1024 GPUs on both Summit and Perlmutter.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Even Higher-Level Synthesis: An Exploration of AI Hardware Accelerators using HLS4ML

With the rise of artificial intelligence, the popularization of deep learning, and a constantly evolving industry, the demand for flexible and efficient tools has never been greater. As algorithms grow more complex, their runtime and energy consumption increase exponentially. Customized hardware accelerators, long used for specific mathematical operations, remain essential for managing modern applications' computational and power demands. Hardware accelerators can speed up complex computations by orders of magnitude, but their manual design and verification processes are often challenging and time-consuming. High-Level Synthesis (HLS) provides a solution by transforming high-level algorithm descriptions, typically written in C++ or SystemC, into synthesizable RTL suitable for hardware implementation. This approach reduces development time for RTL engineers while offering flexibility beyond what traditional handwritten RTL can provide. We extended this capability to the machine-learning domain with the open-source framework hls4ml, which allows neural networks trained in Python frameworks like Tensorflow or PyTorch to be synthesized into efficient hardware representations for the traditional FPGA and ASIC flows. This breakthrough addresses the growing need for reduced design turnaround and easy verification of ML hardware accelerators with low latency and power efficiency constraints. During this tutorial, we will demonstrate how Python complements HLS by simplifying the ML design process, bridging the gap between software and hardware development. Attendees will explore how we translate neural networks modeled in Python into fixed-point C++ models suitable for HLS workflows. We will dive into strategies like Value-Range Analysis and Quantization-Aware Training, which optimize these designs for deployment and evaluate their accuracy, power consumption, and energy efficiency. To exemplify these concepts, experts from Fermilab will share their experiences applying this technology to high-energy physics experiments, where real-time, low-latency processing is critical. Over the years, Fermilab engineers have demonstrated how deep neural networks, optimized for hardware using hls4ml, can meet the stringent requirements of trigger systems at the CERN Large Hadron Collider. These systems rely on rapid decision-making to process immense data volumes while retaining only the most relevant events for further analysis. The application of hls4ml has also been extended to innovative technologies like smart pixel arrays. These smart pixels integrate ML inference capabilities directly into sensor devices, enabling localized data processing at the pixel level. This approach drastically reduces the need to transmit raw data to external processing units, significantly decreasing power consumption and latency. By embedding neural networks within the pixel architecture, the smart pixels can identify and prioritize relevant data in real time, providing a highly efficient solution for edge computing in scenarios such as particle detectors and imaging systems. Fermilab's work highlights the potential of hardware-accelerated ML in scenarios where both speed and power efficiency are mission-critical. Through this tutorial, attendees will gain valuable insights into the challenges and solutions of deploying ML in hardware. Understanding how HLS and hls4ml streamline the development of neural network-based hardware accelerators is fundamental for the industry's future. Participants will learn how these technologies are shaping the future of AI and scientific computing.

Di Guglielmo, Giuseppe [Fermilab]↗